Desertification Investigating And Monitoring With Remote Sensing In Ge Ermu City In Qinghai Province

Currently, desertification is an important environmental problem in the world.Land Desertification is the most serious graveness problem in Qinghai province.Because of the particular environment of geology and climate and human irrationaleconomical activities, Land Desertification in Qinghai Province has the current ofincreasing. So The Investigating and Monitoring to the Land Desertification becomesthe most important problem in Qinghai province. At present, the data of remotesensing has the characteristics of accuracy and timely, it can coverage abroad area. Sowe always use RS information. But the key problem we must solve is to distill thedesertification information from the remote sensing image, and choose the appropriateclassification and appraisement methods. The paper chooses the Ge Ermu City in Qinghai province, and utilizes ETM/TMdata of four period of time. We pick up the desertification information in the remotesensing image, and discuss the classification and appraisement methods. The mainachievement list as following:(1) We analyzed the style of Land Desertification and spectral characteristics ofvegetation and water. Furthermore, we disserted the appraisement method to thestudying of Land Desertification.(2) We bring forward the transformation of MNF, the method of Pixel Purity Indexand the method of n-Dimensioned Visualization. In order to improve the precision ofthe classification result, we combine those methods together to find the purity pixelfrom the image.(3) In the SAM, we use the information of water and vegetation which were obtainedfrom the remote sensing image as the region of interest. The achievement is good. Theprecision of the image classification of the period of the year 2002 is 80.87%. TheKappa Index is 0.77. The precision of the image classification of the year 2000 is87.15%. The Kappa Index is 0.83. The precision of the image classification of theyear 1996 is 66.59%. The Kappa Index is 0.60. The precision of the imageclassification of the year 1994 is 73.92%, The Kappa Index is 0.69.(4)We got the statistical data of Land Desertification in different time, make themean value of two classification images which have the similar achievement time.Through contrast the data of different time, we can see that the area of sand hasincreased 46.14km2, the area of saline has decreased 39.48km2, the area of thevegetation have decreased 61.46km2. The area of sand and saline has added up to6.66km2. The serious situation of Land Desertification keeps increasing. This alsoproves the Land Desertification has been aggravating during the past six years.